Unexpected judgments: the role of gender identity and provocation on blame and affect in a mock jury paradigm

Introduction: Panic defenses are a form of legal defense positing that a defendant is less culpable for their crime because of an extreme emotional reaction. However, limited research has examined the use of panic defenses when the victim is a transgender individual.

Methods: In two studies (Total N = 995) previously existing scales and vignettes were used to evaluate perceptions of panic defenses when the victim was a transgender woman, transgender man, or a gay man. Study 1 (N = 557) used a 2 (sexual provocation or non-sexual provocation) x 3 (transgender woman, transgender man, or gay man victim) design to evaluate perceptions of victim blame, negative affect, and perception that the crime was a hate crime.

Results: Contrary to demographic data trends, Study 1 found that transgender women were rated more positively as victims, were blamed less than transgender men or gay men, and that the crime was more likely to be labeled as a hate crime. Study 2 (N = 438) was a replication of Study 1. While the results were largely non-significant, the trend in means was in the same direction as the findings of Study 1.

Discussion: We argue that these findings, despite not being in line with broader data on the topic, are important evidence that anti-transgender research, particularly in mock jury settings, may need to reimagine existing designs and vignettes to understand rates of violence and discrimination toward transgender communities.

Introduction

In 2002, 17-year-old Gwen Araujo was murdered by four teenage boys based on her identity as a transgender woman. During the trial, the attorneys of two of the defendants argued that learning Araujo's transgender identity provoked the defendants to murder Araujo in the heat of passion, a defense known as the transgender panic defense. The two defendants were convicted of a lesser sentence of second-degree murder and were found not to have committed a hate crime, suggesting the defense may have been partially successful (Lee and Kwan, 2014).

Unfortunately, Araujo's case is not unique. Transgender individuals are more than four times as likely as their cisgender counterparts to experience violent crimes (The Williams Institute, 2021). Although there is no difference in overall rates of physical violence between transgender men and women (The Williams Institute, 2021), transgender women are murdered at higher rates than transgender men (Mallory et al., 2021). Given the prevalence of violence toward both transgender communities, public acceptance of legal defenses such as the transgender panic defense in the United States are particularly important to understand.

LGBTQ+ panic defenses

The transgender panic defense is part of a larger set of legal defenses known as “provocation defenses,” which argue that a defendant lost self-control in response to an extreme emotional reaction (Cheyne and Dennison, 2005). Historically, LGBTQ+ panic defenses have been used to argue that the defendant is less culpable for their crime based on the extreme emotional reaction upon learning the victim's gender or sexual identity (Chen, 2000; Lee and Kwan, 2014; Tomei and Cramer, 2016). In the United States, LGBTQ+ panic defenses are still admissible as a legal defense strategy in 34 states and 5 territories (Movement Advancement Project, 2023).

Although the LGBTQ+ community makes up about 5.6% of the adult United States population, members of the LGBTQ+ community are disproportionately represented in hate crime statistics. In 2019, 19.4% of hate crimes were motivated by discrimination against the victim's gender or sexual identity (Hate Crimes, 2023). The full numbers on frequency of violence against the LGBTQ+ community may be even higher than the figures reported by official sources due to underreporting (Holden, 2020). LGBTQ+ panic defenses provide a legal loophole for anti-LGBTQ+ violence by reducing charges for defendants in 32.7% of cases where the defense was employed (Mallory et al., 2021; Andresen, 2024). Although notable research has examined perceptions of the gay panic defense, research on transgender panic defenses is extremely limited.

Anti-transgender prejudice

Within the LGBTQ+ community, transgender individuals, regardless of their sexual orientation, experience the highest rates of violence and discrimination. Transgender people are more than 2.5 times more likely to experience violence than their cisgender sexual minority counterparts (Truman and Morgan, 2022). Transgender women are particularly likely to experience violence because of their identity; 87% of the murders of transgender individuals evaluated by the Williams Institute were murders of transgender women (Mallory et al., 2021).

Primary emotions of anger and disgust contribute to prejudice toward gender and sexual minorities (Mackie et al., 2000; Huffaker and Kwon, 2016). However, previous research on attitudes about transgender individuals demonstrates that anti-transgender attitudes tend to be more extreme and more negative than attitudes toward cisgender gay men or lesbians (Norton and Herek, 2013). Moreover, attitudes toward transgender women may be more negative compared to transgender men (Totton and Rios, 2021). Negative attitudes may expand to other consequential beliefs as well. Previous research has shown that anti-transgender attitudes also correlate with higher levels of victim blame (Thomas et al., 2016). Transgender individuals, and particularly transgender women, may face a compounding discrimination within the legal system because they are subject to higher rates of violence and more negative attitudes from others, which may in turn impact perceptions of their victimhood.

Previous research on attitudes toward LGBTQ+ panic defenses

Although research on the use of transgender panic defenses is limited, research has explored attitudes about the use of the gay panic defense. Prior research on the gay panic defense highlights acceptance for the use of the defense in certain circumstances. Specifically, findings demonstrate that mock jurors engage in greater levels of victim blaming when the murder took place in a more general setting (a “local” bar rather than a gay bar) and when the victim had made a sexual advance toward the perpetrator (i.e., a sexual provocation; Plumm et al., 2010). Making a sexual advance may be perceived by jurors as taking an action that “facilitated” the crime and thus may lead to greater victim blame or lower defendant blame (Kelley, 1987; Plumm et al., 2010). Moreover, research has demonstrated that political conservatives are less punitive toward defendants when the gay panic defense is employed (Salerno et al., 2015), and that homophobia levels were correlated with a greater acceptance of the use of the gay panic defense (Michalski and Nunez, 2020). Finally, previous research has found that homophobia is related to more lenient sentencing lengths in a gay-panic mock jury study (Kraus and Ragatz, 2011). Taken together, this previous work demonstrates a willingness of participants to accept the use of the gay panic defense as a mitigating factor in certain circumstances (e.g., when the victim makes a sexual advance, when the crime is committed outside of a “queer space,” and amongst conservative and/or homophobic participants).

The current study

The current research evaluates attitudes toward transgender and gay victims in a mock jury paradigm where a panic defense is used. Given that attitudes toward transgender individuals of any sexual orientation tend to be more extreme and more negative than attitudes toward cisgender gay men or lesbians (Norton and Herek, 2013; Totton and Rios, 2021) we anticipate that participants will show greater leniency toward defendants using the transgender panic defense than those using the gay panic defense. Providing some level of support for this, recent research demonstrated that participants exposed to the transgender panic defense were more lenient in their convictions than participants exposed to a control condition (Michalski et al., 2022).

Moreover, since murder rates are notably higher among transgender women than either gay men or transgender men (Holden, 2020; Hate Crimes, 2023), and since both transgender women and men experience higher rates of violence than their cisgender queer counterparts (Mallory et al., 2021; Truman and Morgan, 2022) it is important to understand perceptions of transgender women and transgender men separately as victims. No research to our knowledge has examined differences in acceptance of the transgender panic defense in comparison to the gay panic defense or examined difference in the use of the transgender panic defense with transgender women vs. transgender men. This study uses previously existing scales and adapted vignettes from research on the gay panic defense aims to fill that gap.

Hypotheses

H1: Previous research suggests that for groups considered to be outgroups in society, anger and disgust are primary emotions that may drive negativity against them (Mackie et al., 2000; Taylor, 2007; Huffaker and Kwon, 2016). Given previous research demonstrating that transgender individuals, and particularly transgender women and viewed more negatively than gay men (Norton and Herek, 2013; Totton and Rios, 2021) we hypothesize that mock jurors will show the highest levels of negative affect (as measured by anger and disgust) toward the transgender woman, followed by transgender man, with the lowest levels of negative affect toward the gay man as a victim.

H2: Given previous research demonstrating the role of provocation by Plumm et al. (2010), we anticipate that participants will have more negative affect about the victim when a sexual provocation (rather than a non-sexual provocation) is made. A sexual advance (provocation) may be viewed as “facilitating” the aggressive act, thus increasing negativity toward the victim (Kelley, 1987; Plumm et al., 2010).

H3: Similarly, given that negative attitudes are correlated with victim blame toward transgender individuals (Thomas et al., 2016) we hypothesize that mock jurors will show the highest levels victim blame when the victim is a transgender woman, followed by transgender man, with the lowest levels of victim blame toward the gay man.

H4: Given previous research demonstrates that provocation may be viewed as “facilitating” the aggressive act (Kelley, 1987) and may lead to greater levels of victim blame (Plumm et al., 2010), we anticipate that participants will blame the victim more in the sexual provocation condition than the non-sexual provocation condition.

H5: In reverse of our predictions for the victim, we hypothesize that participants will have the lowest levels of negative affect toward the defendant when the victim is a transgender woman, followed by transgender man, with the highest negative affect toward defendants who murdered a gay man.

H6: Similarly, we hypothesize that participants will view the defendant as less blameworthy for the murder when the victim is a transgender woman, followed by transgender man, with the highest levels defendant blame toward defendants who murdered a gay man.

H7: Considering past findings that participants have higher levels of prejudice toward transgender women in particular (Totton and Rios, 2021) and that participants are less likely to view a crime as a hate crime when they have higher levels of prejudice (Michalski and Nunez, 2020) we hypothesize that participants will be least likely to rate the crime as a hate crime in the transgender woman condition, and most likely to rate the crime as a hate crime in the gay man condition.

H8: Given that previous research has found that sexual prejudice is related to sentencing lengths (Kraus and Ragatz, 2011) and that transgender women experience greater levels of prejudice than transgender men or gay men (Totton and Rios, 2021), we hypothesized that defendants who murdered a gay man would receive the longest sentence length recommendation, followed by transgender men, followed by transgender women.

Although we explored the relationship between conditions (sexual provocation or not) with all variables, as well as the interaction between conditions (gender and provocation conditions) we did not have a-priori hypotheses about the interactions between variables or the relationship between provocation and the additional variables (e.g., sentencing length, hate crime perceptions, or attitudes toward the defendant). Indeed, previous research failed to find a relationship between provocation condition and hate crime perception (Plumm et al., 2010).

Study 1 methods

Participants

Participants were 775 adults residing in the United States (Mean age = 30.9 years old, SD = 10.6 years) recruited through Prolific.com, an online survey website. The initial sample included 329 male participants (42.1%), 444 female participants (56.9%), and 2 non-binary/third gender participants (0.3%). Non-binary/third-gender (n = 2) and transgender (n = 12) participants were excluded from the study. Two-hundred and 18 responses were excluded because participants either did not complete the survey (N = 56) or because they failed the manipulation check (N = 162), leaving 557 participants in the final analysis. Overall, participants rated their political orientation as slightly more liberal (M = 3.64, SD = 1.99) on a scale from 1 (liberal) to 7 (conservative). A sensitivity analysis using G^*Power 3.1 (Faul et al., 2009) demonstrated that this was a sufficient sample for a 2 × 3 design, to detect medium effects (f = 0.25) with 80% power and an alpha of 0.05.

Procedure and materials

First, participants provided demographic information including political orientation, gender identity, sexual orientation, and age. Demographics were asked at the beginning of the study so that transgender and non-binary participants could be excluded. Previous research suggests that placement of demographics does not meaningfully impact responses throughout the survey (Drummond et al., 2008; Teclaw et al., 2012).

Next, all participants read jury instructions based on previous mock jury research (Salerno et al., 2015). These instructions can be viewed in the Supplementary material. Next, participants were randomly assigned to one of six vignettes. After reading their assigned vignette, all participants answered a series of pre-existing scales and questions regarding blame of the defendant and victim, affect toward the victim and defendant, and the extent to which they considered the vignette to be indicative of a hate crime. The vignettes and measure are described below and listed fully in the Supplementary material.

Vignettes

Participants were asked to read one of six randomly assigned vignettes in accordance with the 3 (gender identity: transgender woman, transgender man, gay man) x 2 (sexual provocation or non-sexual provocation) between-subjects design. In each vignette, a cisgender male defendant was provoked, and committed a homicide. The victim, either a transgender woman, transgender man, or a gay man, provoked the defendant either by insulting the defendant's wife and yelling at the defendant or by making a sexual advance on the defendant by placing the defendant's hand on their genitalia. Each of the six vignettes was an adapted version from Salerno et al.'s (2015) study. After completing general questions, participants were asked to indicate the victim's gender or sexual orientation as a manipulation check.

Juror verdict and sentencing

Participants were asked to indicate whether they would convict the defendant of murder (n = 259), manslaughter (n = 292), or neither (n = 6). They were then asked to provide a sentence corresponding to their verdict decision using a sliding scale ranging from 0 to 50 years (50 being a life sentence).

Perceptions of the defendant and the victim blame

Participants then completed a 5-item perception of the defendant and victim blame scale from Michalski and Nunez (2020) [e.g., please indicate the extent to which you think the defendant was violent or non-violent on a scale ranging from 1 (extremely violent) to 7 (extremely non-violent)] (M = 5.58, SD = 1.09, α = 0.82). Participants then answered identical questions about their perception of the victim (M = 3.90, SD = 1.29, α = 0.83).

Negative affect toward the victim and defendant

Next, participants completed a two-item measure from Michalski and Nunez (2020) that assessed participants' anger and disgust toward the actions of the victim on a scale ranging from 1 (not angered/disgusted at all) to 7 (extremely angered/disgusted; M = 3.71, SD = 1.50). These questions were then repeated for the defendant (M = 5.51, SD = 1.37). Correlational analyses indicated that anger and disgust were highly correlated (0.77 for the defendant responses and 0.79 for the victim).

Hate crime perception

Finally, all participants were asked to rate the extent to which they perceived the crime to be indicative of a hate crime in a single item measure (Michalski and Nunez, 2020; rated from 1= strongly disagree; 7 = strongly agree).

All materials for both studies can be found in the Supplementary material.

Study 1 results

Statistical analyses

SPSS version 27 was used to conduct a series of univariate analyses of variance (ANOVA) tests. In each ANOVA, political orientation and participant gender were included as covariates. All interactions were non-significant and are presented in the Supplementary material for brevity.

Negative affect toward the victim

H1: victim gender

There was a significant effect of victim gender on negative affect toward the victim (F_{2, 548} = 8.05, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.029). However, contrary to Hypothesis 1, participants had significantly less negative affect toward transgender women (M = 3.40, SE = 0.10) as victims in comparison to transgender men (M = 3.97, SE = 0.11; p < 0.001) or gay men (M = 3.84, SE = 0.10; p = 0.003). There was no difference between negative affect toward transgender men and gay men (p = 0.386).

H2: provocation

In line with Hypothesis 2, participants had more negative affect toward the victim in the sexual provocation condition (M = 4.12, SE = 0.08) than the non-sexual provocation (M = 3.36, SE = 0.09; F_{2, 548} = 38.84, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.066).

Perceptions of victim blame scale

H3: victim gender

Victim gender had a significant effect on participants' levels of victim blame (F_{2, 548} = 8.15, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.029). However, against Hypothesis 3, participants blamed the victim less when the victim was a transgender woman (M = 3.61, SE = 0.72) than when the victim was a transgender man (M = 4.00, SE = 0.08; p < 0.001), or a gay man (M = 3.93, SE = 0.07; p = 0.002). There were no significant differences in defendant blame when the victim was a transgender woman or a gay man (p = 0.481).

H4: provocation

In line with Hypothesis 4, participants blamed the victim more in the sexual provocation condition (M = 4.10, SE = 0.06) than the non-sexual provocation (M = 3.59, SE = 0.06; F_{2, 548} = 35.43, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.061).

Negative affect toward the defendant

H5: victim gender

There was a significant effect of victim gender on negative affect toward the defendant (F_{2, 548} = 4.65, p = 0.010, ${\eta }_{\text{p}}^2$ = 0.017). However, opposed to H5, participants had more negative affect toward defendants who murdered a transgender woman (M = 5.74, SE = 0.09) in comparison to a transgender man (M = 5.33, SE = 0.10; p = 0.003) or a gay man (M = 5.41, SE = 0.09; p = 0.046). There was no difference between negative affect toward defendants who murdered a transgender or gay man (p = 0.293).

Provocation

There was no significant difference in negative affect toward the defendant between the sexual provocation condition (M = 5.57, SE = 0.08) and the non-sexual provocation (M = 5.46, SE = 0.08; F_{2, 548} = 0.98, p = 0.323, ${\eta }_{\text{p}}^2$ = 0.002).

Perceptions of defendant blame scale

H6: victim gender

Victim gender had a significant effect on participants' levels of defendant blame (F_{2, 548} = 3.89, p = 0.021, ${\eta }_{\text{p}}^2$ = 0.014). In opposition to Hypothesis 6, participants blamed the defendant less when the victim was a transgender man (M = 5.01, SE = 0.70) than when the victim was a transgender woman (M = 5.28, SE = 0.07; p = 0.006), but not more than when the victim was a gay man (M = 5.13, SE = 0.07; p = 0.111). There were no significant differences in defendant blame when the victim was a transgender woman or a gay man (p = 0.217).

Provocation

The provocation condition did not have a significant effect on participants' levels of defendant blame (F_{1, 548} = 0.76, p = 0.385, ${\eta }_{\text{p}}^2$ = 0.001 (Sexual advance provocation M = 5.18, SE = 0.06; non-sexual provocation M = 5.11, SE = 0.05).

Hate crime perception

H7: victim gender

There was a significant effect of victim gender on sentencing length (F_{2, 548} = 4.80, p = 0.009, ${\eta }_{\text{p}}^2$ = 0.017). However, contrary to Hypothesis 7, participants were more likely to see the crime as a hate crime when the victim was a transgender woman (M = 4.09, SE = 0.13) than a gay man (M = 3.54, SE = 0.13; p = 0.002). There was no difference in perception of hate crimes for a transgender man (M = 3.82, SE = 0.14) with either a transgender woman (p = 0.145) or gay man (p = 0.127).

Provocation

Participants were more likely to rate the crime as a hate crime in the sexual provocation condition (M = 4.08, SE = 0.12) than the non-sexual provocation (M = 2.14, SE = 0.12; F_{2, 548} = 144.56, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.208).

Sentencing severity

H8: victim gender

There was a significant effect of victim gender on sentencing length (F_{2, 548} = 7.37, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.026). However, contrary to Hypothesis 8, participants recommended longer sentences for defendants who murdered a transgender woman (M = 26.66, SE = 0.96) than either a transgender man (M = 22.25, SE = 1.03; p = 0.002) or gay man (M = 21.99, SE = 0.96; p < 0.001). There was no difference between sentencing recommendations toward defendants when the victim was a transgender or gay man (p = 0.855).

Provocation

Participants were marginally more likely to recommend a longer sentence in the sexual provocation condition (M = 24.75, SE = 0.83) than the non-sexual provocation (M = 22.52, SE = 0.78; F_{2, 548} = 3.86, p = 0.050, ${\eta }_{\text{p}}^2$ = 0.007).

Study 1 discussion

Contrary to our hypotheses, participants rated the defendant as being more blame worthy, had a lower opinion of the defendant, recommended a longer sentence, and were more likely to view the crime as a hate crime when the victim was a transgender woman. Similarly, participants blamed the victim the least and had the least negative affect toward the transgender woman victim. These finding do not align with real world data demonstrating that transgender women face higher rates of violence and murder in comparison to gay men (Truman and Morgan, 2022), as well as previous research suggesting that transgender women are rated more negatively than gay men or transgender men (Norton and Herek, 2013; Totton and Rios, 2021). Although Hypotheses 1, 3, and 4–8 were not supported, we did find support for Hypotheses 2 and 4. Participants rated victims more negatively and blamed them more when the defendant was provoked. Study 2 attempted to replicate the findings of Study 1 and included a scale previously used to evaluate negative attitudes toward transgender women, transgender men, and gay men. These questions were based on Totton and Rios (2021) which demonstrated that transgender individuals were viewed as more deceptive than gay men, and that this effect was particualrly pronounced for transgender women. This measure was added as an opportunity to further assess participant attitudes using a previously validated scale. Despite the unexpected results of Study 1, we maintained the hypotheses of Study 1 given data suggesting that transgender individuals are rated more negatively in other settings (e.g., Norton and Herek, 2013; Totton and Rios, 2021) and face higher rates of violence (Mallory et al., 2021; The Williams Institute, 2021), but a final hypothesis was added based on previous literature.

H9: Based on research by Totton and Rios (2021), Transgender women will be rated as more deceptive than transgender men or gay men.

Study 2 methods

Participants

In Study 2, participants were once again recruited through Prolific. Participants who participated in Study 1 were ineligible to participate. Participants were immediately excluded (N = 52) if they did not correctly identify the gender identity or sexual orientation of the victim in the vignette to ensure attentiveness. Four-hundred and thirty-eight cisgender adults residing in the United States (Mean age = 41.33 years old, SD = 14.45) completed the study. The initial sample included 211 male participants (48.2%) and 227 female participants (51.8%). Overall, participants rated their political orientation as slightly more liberal (M = 3.51, SD = 1.71; 1 = very liberal, 7 = very conservative). A sensitivity analysis using G^*Power 3.1 (Faul et al., 2009) demonstrated that this was a sufficient sample for a 2 × 3 design, to detect medium effects (f = 0.25) with 80% power and an alpha of 0.05.

Procedure and materials

Procedure and materials were identical to Study 1, with one addition. Participants were asked to complete a four itemperceived deception scale about the marginalized identity of the victim (e.g., “It is dishonest for a transgender woman/transgender man/gay man to not reveal their identity to others”; overall: M = 3.06, SD = 1.68; transgender women α −0.977; transgender men α = 0.97; gay men α = 0.96) based on Totton and Rios (2021). Scale means and standard deviations are also presented in the Supplementary material.

Study 2 results

Statistical analyses

All analyses were identical to Study 1.

Negative affect toward the victim

H1: victim gender

There was no significant effect of victim gender on negative affect toward the victim (F_{2, 430} = 0.23, p = 0.793, ${\eta }_{\text{p}}^2$ = 0.001).

H2: provocation

In line with Hypothesis 2 and Study 1, participants had more negative affect toward the victim in the sexual provocation condition (M = 3.99, SE = 0.10) than the non-sexual provocation (M = 3.20, SE = 0.11; F_{2, 430} = 26.76, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.059).

Perceptions of victim blame scale

H3: victim gender

Victim gender did not have a significant effect on participants' levels of victim blame (F_{2, 438} = 1.71, p = 0.183, ${\eta }_{\text{p}}^2$ = 0.008; Transgender woman M = 4.01, SE = 0.10; Transgender man M = 4.23, SE = 0.10; Gay man M = 4.22, SE = 0.09).

H4: provocation

In line with Hypothesis 4 and with Study 1, participants blamed the victim more in the sexual provocation condition (M = 4.27, SE = 0.07) than the non-sexual provocation (M = 4.03, SE = 0.08; F_{2, 438} = 5.16, p = 0.024, ${\eta }_{\text{p}}^2$ = 0.012).

Negative affect toward the defendant

H5: victim gender

There was a marginal but non-significant effect of victim gender on negative affect toward the defendant (F_{2, 438} = 2.69, p = 0.069, ${\eta }_{\text{p}}^2$ = 0.012). Although the effect was non-significant, the marginal pattern of means followed a similar pattern to Study 1, wherein participants had the most negative attitudes toward defendants who murdered a transgender woman (M = 5.53, SE = 0.13), followed by a gay man (M = 5.37, SE = 0.11), and a transgender man (M = 5.11, SE = 0.13).

Provocation

There was no significant difference in negative affect toward the defendant in the sexual provocation condition (M = 5.41, SE = 0.10) than the non-sexual provocation (M = 5.30, SE = 0.11; F_{2, 438} = 1.138, p = 0.287, ${\eta }_{\text{p}}^2$ = 0.003).

Perceptions of defendant blame scale

H6: victim gender

Victim gender did not significantly affect participants' levels of defendant blame (F_{2, 438} = 1.14, p = 0.320, ${\eta }_{\text{p}}^2$ = 0.004; Transgender woman M = 5.97, SE = 0.07; Transgender man M = 5.83, SE = 0.07; Gay man M = 5.96, SE = 0.06).

Provocation type

Similar to Study 1, the provocation condition did not have a significant effect on participants' levels of defendant blame (F_{1, 438} = 1.79, p = 0.182, ${\eta }_{\text{p}}^2$ = 0.005; Sexual advance provocation M = 5.87, SE = 0.06; non-sexual provocation M = 5.97, SE = 0.06).

Hate crime perception

H7: victim gender

There was not a significant effect of victim gender on sentencing length (F_{2, 430} = 1.42, p = 0.224, ${\eta }_{\text{p}}^2$ = 0.007; transgender woman M = 3.31, SE = 0.15; transgender man M = 3.04, SE = 0.15); gay man M = 2.98, SE = 0.13).

Provocation

Participants were more likely to rate the crime as a hate crime in the sexual provocation condition (M = 4.08, SE = 0.12) than the non-sexual provocation (M = 2.14, SE = 0.12; F_{2, 430} = 132.68, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.236).

Sentencing severity

H8: victim gender

Unlike Study 1, there was no significant effect of victim gender on sentencing length (F_{2, 430} = 1.46, p = 0.230, ${\eta }_{\text{p}}^2$ = 0.007). However, the trend in means was similar to Study 1 with defendants who murdered a transgender woman (M = 23.12, SE = 1.21) being recommended for a non-significantly longer sentence than either a transgender man (M = 20.67, SE = 1.20) or gay man (M = 20.61, SE = 1.07).

Provocation

There was no difference in sentence recommendation between the sexual provocation condition (M = 21.49, SE = 0.98) and the non-sexual provocation (M = 21.44, SE = 0.92; F_{2, 430} = 0.232, p = 0.630, ${\eta }_{\text{p}}^2$ = 0.001).

Perceived deception scale

H9: victim gender

Victim gender had a significant effect on participants' perception that the victim was being deceptive (F_{2, 430} = 32.06, p < 0.001, ${\eta }_{\text{p}}^2$ = 0.130). In line with Hypothesis 9 and previous research, but not in line with other findings in this paper, participants believed that transgender women were more deceptive (M = 3.55, SE = 0.12) than gay men (M = 2.43, SE = 0.11; p < 0.001). Similarly, transgender men (M = 3.47, SD = 0.12) were also rated as more deceptive than gay men (p < 0.001). There were no significant differences in perceived deceptiveness of transgender women or transgender men (p = 0.620).

Provocation

There were no significant differences in perceptions of deceptiveness based on provocation condition (Sexual advance: M = 4.10, SE = 0.06; non-sexual provocation: M = 3.59, SE = 0.06; F_{2, 430} = 0.721, p = 0.396, ${\eta }_{\text{p}}^2$ = 0.002).

Study 2 discussion

The results of Study 2 do not provide full support for the initial hypotheses nor the results of Study 1. Participants demonstrated only marginally more negative affect toward the defendant when the victim was a transgender woman rather than a transgender man or gay man. However, there were no significant differences in victim blame, defendant blame, or negative affect toward the victim based on victim gender as was found in Study 1. However, Study 2 directly replicated Study 1 in support for Hypotheses 2 and 4, with victims being blamed more and eliciting more negative affect when the defendant was provoked through a sexual advance. Finally, transgender women and men were both perceived as more deceptive than gay men, a finding that is partially in line with Hypothesis 9 and previous research which found that transgender women were perceived as more deceptive than either transgender men or gay men (Totton and Rios, 2021). Despite obtaining an appropriate sample size to detect medium effects, it is possible that this study was underpowered for the effect sizes observed in Study 1.

General discussion

This research examined perceptions of the transgender panic defense. Based on demographic data surrounding hate crimes (e.g., Mallory et al., 2021; Truman and Morgan, 2022), as well as previous research demonstrating that attitudes are more negative toward transgender people (particularly transgender women) than gay men (Norton and Herek, 2013; Totton and Rios, 2021), we hypothesized that participants would have the most blame and negative attitudes toward transgender women as victims and would have the least blame and most positive attitudes toward defendants who murdered a transgender woman. However, our results found no support for this. Indeed, in Study 1, we found that participants had the least blame and negative attitudes toward transgender women as victims and showed the greatest negativity, blame, and belief that the crime was a hate crime toward defendants who murdered a transgender woman. While non-significant, the trend of means in Study 2 generally also supported this pattern.

These unexpected findings may be evidence of participant response monitoring. Based on media representations or general societal discourse, participants may have recognized the societal challenges faced by transgender women and responded with greater positivity toward trans women as victims to go against societal discrimination. However, this explanation and the findings in this paper do not align with Michalski et al.'s (2022) work on the transgender panic defense specifically, nor with previous work on attitudes about transgender women in relation to transgender men or gay men (Norton and Herek, 2013; Totton and Rios, 2021). Moreover, participants did report greater perceived deceptiveness of transgender women, suggesting they were not altering their attitudes on those questions, and supporting results from previous research studies (Totton and Rios, 2021; Totton et al., 2023).

Alternatively, it is possible that participants were more negative toward transgender men and gay men as a biproduct of the current sociopolitical landscape in the United States where there is has been a recent uptick of anti-LGBTQ+ (and particularly anti-trans) hate groups and legislation (American Civil Liberties Union, 2024; Southern Poverty Law Center, 2024). Spillover effects from the rhetoric shared in these hate groups or legislation may have impacted participant views of transgender folks or cisgender gay men in distinct ways. For example, anti-transgender legislation or anti-transgender rhetoric often refers to transgender people using incorrect gender identifiers. This language may lead to confusion amongst participants about the gender identity of the victim (despite including a manipulation check in Study 2). Similarly, the inclusion of sexual orientation information in the gay man but not transgender man or transgender woman conditions may have led participants to overly emphasized the victim's sexual orientation, and thus to hyper-sexualize the victim in the gay male condition. This potential issue could be resolved through a replication including additional clarifying questions about the victim's identity to ensure participant understanding. Moreover, follow up studies should include sexual orientation information about transgender victims as well as other cisgender sexual minority conditions (i.e., lesbians) to account for potential sexualization effects.

Alternatively, the results may have been due to the manipulation used. While the manipulation was based directly on a manipulation used successfully in an evaluation of the gay panic defense (Salerno et al., 2015), the vignette may have elicited stronger responses in relation to homophobic attitudes. This would explain the general trend in the data toward more negative attitudes toward both transgender and gay men over transgender women. Since the vignette in the current study had some level of similarity to real world instances of the use of the transgender panic defense, further research should evaluate participant responses using a variety of potential vignettes to understand the role of the vignette in the current findings. For example, the current studies do not include Lesbian women as a victim condition. Inclusion of this identity may provide a meaningful comparison group to understand the role of gender identity.

Furthermore, future research should evaluate potential moderating effects that might explain some of the relationships observed. Namely, future research would benefit from explicitly measuring pre-existing prejudice levels of participants and looking for potential moderating effects based on preconceived notions about transgender individuals or gay men. While the present study included political orientation as a covariate, treating it as a moderator might yield interesting results.

Finally, it is possible that participants were more negative toward transgender men and gay men than transgender women as victims. While data suggests that the transgender panic defense is used more frequently when the victim is a transgender woman (Mallory et al., 2021) and that transgender women elicit overall more negative attitudes than either transgender men or gay men (Totton and Rios, 2021), it is possible that these effects do not carry over to perceptions of victimhood, in particular. Previous research has found that harassment of transgender women is viewed differently than harassment of cisgender heterosexual or lesbian women (Mezzapelle and Reiman, 2022). Although this research did not explore attitudes toward transgender men or gay men, it does suggest that transgender women may face disparate victimhood perceptions for the same offense. Yet, this explanation is not in line with previous research demonstrating that anti-transgender prejudice is related to victim blame either (Thomas et al., 2016). However, this is the first paper to our knowledge that compares attitudes toward transgender men, transgender women, and gay men in a mock jury paradigm, making the results challenging to directly compare to previous research.

Importantly, we believe it is critical that the findings of this paper not be used to imply that transgender women face less stigma in the legal system than their transgender or gay male counterparts. Rather, we suggest that taken together, the findings of this paper point to a need for more comprehensive research on attitudes surrounding transgender individuals, particularly in jury settings. Given the prolific rise in anti-transgender legislation (Lavietes and Ramos, 2022; Human Rights Campaign, 2023), and the continued acceptance of the transgender panic defense in most US states (Movement Advancement Project, 2023) better understanding of attitudes and treatment of transgender people within the legal system is critical.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by Amherst College Institutional Review Board. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

RT: Conceptualization, Formal analysis, Methodology, Project administration, Writing – original draft, Writing – review & editing. KM: Conceptualization, Methodology, Writing – original draft, Writing – review & editing.

Funding

The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frsps.2024.1327799/full#supplementary-material

References

American Civil Liberties Union (2024). Mapping Attacks on LGBTQ Rights in U.S. State Legislatures in 2024. Available online at: https://www.aclu.org/legislative-attacks-on-lgbtq-rights-2024 (accessed July 1, 2024).

Google Scholar

Andresen, W. C. (2024). I Track Murder Cases That Use the ‘Gay Panic Defense,' a Controversial Practice Banned in 9 States. https://theconversation.com/i-track-murder-cases-that-use-the-gay-panic-defense-a-controversial-practice-banned-in-9-states-12

Google Scholar

Chen, C. (2000). Provocation's privileged desire: the provocation doctrine, homosexual panic, and the non-violent unwanted sexual advance defense. Cornell J. Law Publ. Pol. 10, 195–236.

Google Scholar

Cheyne, N., and Dennison, S. (2005). An examination of potential reform to the provocation defense: the impact of gender of the defendant and the suddenness requirement. Psychiat. Psychol. Law 12, 388–400. doi: 10.1375/pplt.12.2.388

Crossref Full Text | Google Scholar

Drummond, F. J., Sharp, L., Carsin, A., Kelleher, T., and Comber, H. (2008). Questionnaire order significantly increased response to a postal survey sent to primary care physicians. J. Epidemiol. 61, 177–185. doi: 10.1016/j.jclinepi.2007.04.012

PubMed Abstract | Crossref Full Text | Google Scholar

Faul, F., Erdfelder, E., Buchner, A., and Lang, A. G. (2009). Statistical power analyses using G^* Power 3.1: tests for correlation and regression analyses. Behav. Res. Methods 41, 1149–1160. doi: 10.3758/BRM.41.4.1149

PubMed Abstract | Crossref Full Text | Google Scholar

Hate Crimes (2023). 2019 Hate Crimes Statistics. Available online at: https://www.justice.gov/hatecrimes/2019-hate-crime-statistics

Google Scholar

Holden, A. (2020). The Gay/Trans Panic Defense: What It Is, and How to End It. American Bar Association. Available online at: https://www.americanbar.org/groups/crsj/publications/member-features/gay-trans-panic-defense/ (accessed September 28, 2023).

Google Scholar

Huffaker, L., and Kwon, P. (2016). A comprehensive approach to sexual and transgender prejudice. J. Gay Lesbian Soc. Serv. 28, 195–213. doi: 10.1080/10538720.2016.1191405

Crossref Full Text | Google Scholar

Human Rights Campaign (2023). Human Rights Campaign Foundation State Equality Index: 91% of Anti-LGBTQ+ Bills in 2022 Failed to Become Law. Human Rights Campaign. Available online at: https://www.hrc.org/press-releases/human-rights-campaign-foundation-stateequality-index-91-of-anti-lgbtq-bills-in-2022-failed-to-become-law (accessed September 15, 2023).

Google Scholar

Kelley, H. H. (1987). “Causal schemata and the attribution process,” in Preparation of This Paper Grew Out of a Workshop on Attribution theory held at University of California (Los Angeles, CA: Lawrence Erlbaum Associates, Inc).

Google Scholar

Kraus, S. W., and Ragatz, L. L. (2011). Gender, jury instructions and homophobia: what influence do these factors have on legal decision making in a homicide case where the defendant utilized the homosexual panic defense? Crim. Law Bullet. 47, 237–256.

Google Scholar

Lavietes, M., and Ramos, E. (2022). Nearly 240 Anti-LGBTQ Bills Fled in 2022 so Far, Most of Them Targeting Trans People. NBC News. Available online at: https://www.nbcnews.com/nbc-out/out-politicsand-policy/nearly-240-anti-lgbtq-bills-fled-2022-far-targetingtrans-people-rcna20418 (accessed September 15, 2023).

Google Scholar

Lee, C., and Kwan, P. (2014). The trans panic defense: masculinity, heteronormativity, and the murder of transgender women. Hastings Law J. 66, 77–132. doi: 10.2139/ssrn.2430390

Crossref Full Text | Google Scholar

Mackie, D. M., Devos, T., and Smith, E. R. (2000). Intergroup emotions: explaining offensive action tendencies in an intergroup context. J. Personal. Soc. Psychol. 79, 602–616. doi: 10.1037/0022-3514.79.4.602

PubMed Abstract | Crossref Full Text | Google Scholar

Mallory, C., Sears, B., and & Vasquez, L. A. (2021). Banning the Use of Gay and Trans Panic Defenses. UCLA School of Law: Williams Institute. Available online at: https://williamsinstitute.law.ucla.edu/publications/model-leg-gay-trans-panic/ (accessed September 1, 2023).

Google Scholar

Mezzapelle, J. L., and Reiman, A. K. (2022). How do people perceive sexual harassment targeting transgender women, lesbians, and straight cisgender women? J. Exp. Psychol. Appl. 28:644. doi: 10.1037/xap0000361

PubMed Abstract | Crossref Full Text | Google Scholar

Michalski, N. D., Bitter, A. N., and Nuñez, N. (2022). “Trans folks are in the crosshairs”: jury decision-making and the trans panic defense. J. Interpers. Viol. 37, 22453–22474. doi: 10.1177/08862605211072165

PubMed Abstract | Crossref Full Text | Google Scholar

Michalski, N. D., and Nunez, N. (2020). When is “gay panic” accepted? Exploring juror characteristics and case type as predictors of a successful gay panic defense. J. Interpers. Viol. 2020:886260520912595. doi: 10.1177/0886260520912595

PubMed Abstract | Crossref Full Text | Google Scholar

Movement Advancement Project (2023). Equality Maps: Panic Defense Bans. Available online at: https://www.lgbtmap.org/equality-maps/panic_defense_bans (accessed September 27, 2023).

Google Scholar

Norton, A. T., and Herek, G. M. (2013). Heterosexuals' attitudes toward transgender people: Findings from a national probability sample of US adults. Sex Roles 68, 738–753. doi: 10.1007/s11199-011-0110-6

Crossref Full Text | Google Scholar

Plumm, K. M., Terrance, C. A., Henderson, V. R., and Ellingson, H. (2010). Victim blame in a hate crime motivated by sexual orientation. J. Homosexual. 57, 267–286. doi: 10.1080/00918360903489101

PubMed Abstract | Crossref Full Text | Google Scholar

Salerno, J. M., Najdowski, C. J., Bottoms, B. L., Harrington, E., Kemner, G., and Dave, R. (2015). Excusing murder? Conservative jurors' acceptance of the gay-panic defense. Psychol. Publ. Pol. Law 21, 24–34. doi: 10.1037/law0000024

Crossref Full Text | Google Scholar

Southern Poverty Law Center (2024). Anti-LGBTQ. Available online at: https://www.splcenter.org/fighting-hate/extremist-files/ideology/anti-lgbtq (accessed July 1, 2024).

Google Scholar

Taylor, K. (2007). Disgust is a factor in extreme prejudice. Br. J. Soc. Psychol. 46, 597–617. doi: 10.1348/014466606X156546

PubMed Abstract | Crossref Full Text | Google Scholar

Teclaw, R., Price, M. C., and Osatuke, K. (2012). Demographic question placement: effect on item response rates and means of a veterans health administration survey. J. Bus. Psychol. 27, 281–290. doi: 10.1007/s10869-011-9249-y

Crossref Full Text | Google Scholar

The Williams Institute (2021). Transgender People Over Four Times More Likely Than Cisgender People to Be Victims of Violent Crime. Available online at: https://williamsinstitute.law.ucla.edu/press/ncvs-trans-press-release/ (accessed August 24, 2023).

Google Scholar

Thomas, D. M., Amburgey, J., and Ellis, L. (2016). Anti-transgender prejudice mediates the association of just world beliefs and victim blame attribution. Int. J. Transgend. 17, 176–184. doi: 10.1080/15532739.2016.1232627

Crossref Full Text | Google Scholar

Tomei, J., and Cramer, R. J. (2016). Legal policies in conflict: the gay panic defense and hate crime legislation. J. For. Psychol. Pract. 16, 217–235. doi: 10.1080/15228932.2016.1192331

Crossref Full Text | Google Scholar

Totton, R., and Rios, K. (2021). Predictors of anti-transgender attitudes: identity-confusion and deception as aspects of distrust. Self Identity 20, 496–514. doi: 10.1080/15298868.2019.1621928

Crossref Full Text | Google Scholar

Totton, R. R., Rios, K., and Shogren, N. (2023). Distrusted disclosures: deception drives anti-transgender but not anti-atheist prejudice. Front. Psychol. 13:8427. doi: 10.3389/fpsyg.2022.1006107

PubMed Abstract | Crossref Full Text | Google Scholar

Truman, J., and Morgan, R. (2022). Violent Victimization by Sexual Orientation and Gender Identity, 2017–2020. Bureau of Justice Statistics. Available online at: https://bjs.ojp.gov/library/publications/violent-victimization-sexual-orientation-and-gender-identity-2017-2020#additional-details-0 (accessed August 28. 2023).

Google Scholar